© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 18-TRANSCODER-HP
    7. 90010
    Prev
    Next
    INDEX
    Explanations

    improvements and changes

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_18.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.18.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    é«ĺå³°æľŁ
    -0.25
    æĺİæľĿ
    -0.25
    å¼Ģåı£
    -0.25
    rm
    -0.24
    دد
    -0.24
    summ
    -0.24
    æĻļä¸Ĭ
    -0.24
     radios
    -0.24
    sym
    -0.24
    ãĢģãĢĮ
    -0.24
    POSITIVE LOGITS
    uto
    0.28
    ForObject
    0.28
     hers
    0.27
    åĢĴ
    0.27
     cor
    0.25
    egov
    0.25
    afil
    0.24
    Trou
    0.24
     revers
    0.24
    $output
    0.24
    Activations Density 2.782%

    No Known Activations